Remini Web
Use Remini Web with a human-in-the-loop browser automation flow that keeps account login and verification manual.
Quick Start
- Run script help first:
python3 scripts/remini_apply.py --help - Install prerequisites if needed:
python3 -m pip install playwright python3 -m playwright install chromium - Run a standard enhancement:
python3 scripts/remini_apply.py /path/to/photo.jpg
The script opens Chromium, reuses a local browser profile, attempts upload/start actions, and waits for the Remini download.
Core Workflow
- Confirm the input is a local image (
.jpg,.jpeg,.png,.webp,.heic,.heif). - Launch
scripts/remini_apply.pywith a persistent profile. - Log in manually in the browser if needed.
- Let the script try auto-upload and auto-start.
- Complete actions manually in-browser if UI labels differ.
- Click Download in Remini and let the script capture/save the output.
Common Commands
# Standard run (visible browser)
python3 scripts/remini_apply.py ./input/portrait.jpg
# Save to explicit output path
python3 scripts/remini_apply.py ./input/portrait.jpg --output ./output/portrait-remini.png
# Disable action click automation if the UI text changed
python3 scripts/remini_apply.py ./input/portrait.jpg --no-auto-start
# Add a custom action label for one-off UI variants
python3 scripts/remini_apply.py ./input/portrait.jpg --action-label "Create"
# Batch from a shell loop (one process per image)
for img in ./batch/*.{jpg,jpeg,png}; do
python3 scripts/remini_apply.py "$img" --downloads-dir ./output/downloads
doneOperational Rules
- Keep login, MFA, and checkout actions manual.
- Do not store passwords or one-time codes in files.
- Prefer
--no-auto-startwhen Remini changes button text or flow. - Increase
--timeoutfor long-running jobs. - Load troubleshooting notes from
references/workflow.mdwhen upload/download capture fails.
Resources
scripts/remini_apply.py: Interactive Playwright helper for Remini Web.references/workflow.md: UI matching strategy and troubleshooting.